peter@ficc.ferranti.com (peter da silva) (08/19/90)
This discussion does not belong in news.groups. I would suggest comp.std.unix or comp.std.misc. As such, I am directing followups there. After the usual exchange of insults between System V and BSD folks: > Wrong. The term is "standardize existing _practice_". When there is > a conflict in existing practice, the committee's clear responsibility > is to choose the technically superior mechanism. The fourteen character > filename limit that breaks software distribution mechanisms across the > net is _not_ it. If the purpose of POSIX was to chart the future of UNIX, I would agree. POSIX is, at least to begin with, intended to provide a standard for programs intended to run on UNIX-like systems. Don't think of POSIX as requiring that all systems have 14 character file names, but that all programs written for POSIX should be able to assume that 14 character file names exist. > I participated in the net discussions on this standard until the clear > intent of the committee to cater to the least capable _hardware_ at the > expense of C programmers for a decade became overwhelmingly evident. Sigh. You can always write your code for GCC and get all the whiz-bang features you want. But then your program will not be maximally portable. This just reflects the real world. The standard could have included a requirement for 64 or 48 or 36 or 60 bit arithmetic, too. Not all compromise works against you. > >Guess what: most unices have a 14-character filename limit. > From a poster at SCO, that might look like a provident thing to say. I'm not at SCO. We have System III, System V, BSD, and even a V7 system here. We also run a lot of NON-UNIX systems, from CP/M and MS-DOS to VAX/VMS and Exec/8. Guess what: most UNIX systems out there are running Xenix 2.x or System V.3 on 80386 processors with an AT bus. These machines have a 14-character filename limit. I can imagine how you'd feel if POSIX specified, oh, file version numbers. A lot of people consider that a file system that doesn't provide version numbers for files is irretreivably brain dead. Or suppose they specified COFF or ELF link formats... after all, most UNIX systems use them... Not all compromises worked against you. -- Peter da Silva. `-_-' +1 713 274 5180. 'U` peter@ferranti.com (currently not working) peter@hackercorp.com
sakkinen@tukki.jyu.fi (Markku Sakkinen) (08/20/90)
In article <76B5SDD@ggpc2.ferranti.com> peter@ficc.ferranti.com (peter da silva) writes: > ... >I can imagine how you'd feel if POSIX specified, oh, file version numbers. A >lot of people consider that a file system that doesn't provide version numbers >for files is irretreivably brain dead. [...] > ... Right on the nail! I have been working in UNIX environments for 5 years now, and still keep yearning for versions, like VAX/VMS and even RSX-11M had (and have) them. Markku Sakkinen Department of Computer Science University of Jyvaskyla (a's with umlauts) Seminaarinkatu 15 SF-40100 Jyvaskyla (umlauts again) Finland SAKKINEN@FINJYU.bitnet (alternative network address)